Melvyn G. Fishel 10027...@compuserve.com
The Everready Product Selector guide gives some numbers based
on very low current drain and low threshhold voltage (0.9 V/cell):
AA alkaline 2500 mAH
AAA alkaline 1100 mAH
AA Lithium 2600 maH
It's interesting that at low drains, lithiums aren't that much better
than alkaline. My experience has been that at high current drains,
they're a LOT better, like at least 3x, and their cell voltage
stays reasonable for most of the discharge curve. A lot of electronic
equipment is IMHO poorly designed, and poops out when the battery
voltage gets below 1.2 volts per cell or so. Alkalines still
have a lot of energy left when they reach that voltage but many
palmtops etc. can't use this energy. Lithiums stay above 1.2
volts all the way to the end.
There is a Battery Engineering Handbook available from Everready
that has some more detailed graphs and figures, but they're not that
much better. I no longer have a copy though, and my old copy would
be way out of date by now even if I still had it. You could contact
Everready or any other battery vendor to ask for data though.
If you make any of your own measurements, let me know. I've been
meaning to do this (I even have a logging voltmeter that is ideal
for the purpose) but I haven't gotten around to it. I'll post
my results if and when I do some experiments.
MN1500 AA : 2450 mAh
MN2400 AAA : 1120 mAh
... from www.duracell.com, without dependency on discharge current.
Best regards, Jiri
--
+---------------------------------------------------------------+
| Jiri Sedlacek, Information Technology Dept. Policy & Strategy |
| SPT TELECOM a.s., Olsanska 5, 130 00 Prague 3, Czech Republic |
| Phone: +42-2-67142104, Fax: +42-2-6919151, E-mail: j...@spt.cz |
+---------------------------------------------------------------+
the design issue is around the complexity and energy loss
through the DC-DC converter vs. a regulator. a converter which
can handle an input voltage down to 1.8 (.9v * 2 cells)
and output enough current at a stable 2.4+ v will run at
an efficiency of (say) 85%. so, 15% of the battery energy is
lost. it also costs $$ to design and reliably manufacture
such a converter. on the other hand, designing the circuitry
to run with power from a simple regulator down to 2.4v
is much simpler and more efficient, with the downside being
not milking the batteries for all their available power.
so, as with most things, it's not necessarily poor
design as much as the design constraints implied by the target
market - mainly unit cost.
1. In the HP 100/200lx and Palm Pilot, such a converter is already
present. Similarly for my Garmin GPS-45 GPS receiver. Yet the HP
and Garmin can't run below 1.2V/cell or so (I don't know about the Pilot).
Explanation? Note: if the explanation involves the 0.7 volt drop
across silicon PN junctions, keep in mind that the Garmin uses four
AA cells and not two. Also, you'd think these converters could
use power MOSFETS instead of bipolar transistors.
2. DC/DC converters are not the bulky, expensive contraptions they
used to be. Take apart a Pilot and see what I mean. There's an
integrated converter chip plus a few fairly cheap external passive
components and that's it.
3. Many devices that can't afford dc/dc converters don't need regulated
power at all, e.g. they use CMOS logic that can run from 2.5 to 15 volts.
So a simple solution for many of them is to use 3 or 4 cells instead of 2.
Obviously, not all devices have enough space for this, but many do
or could, or they could maybe use a 9 volt battery instead of two AA's.
In equipment designed to be usable with nicad cells, the voltage
warning and shutoff will often be higher than that necessary for
alkalines, to give users enough warning to save their work, given the
rapid voltage drop off of nicads at the end of life. One of these
days I want to read in one of the charts and approximate it, but
typically something like less than five percent of a nicads life is
left at 1.1 volts or so.
--
Richard Dell